The Myers Briggs Type Inventory Charles C. Healy Professor of Education, UCLA In press with Kapes, J.T. et. al. (2001) A counselor s guide to Career Assessment Instruments. (4th Ed.) Alexandria, VA: National Career Development Association Description A revised, improved version of the Myers-Briggs Type Indicator (MBTI) and its Manual were released in 1998. The MBTI continues to focus on selected Jungian (1971) concepts as modified by its authors to gauge people s preferred ways of perceiving, judging, and directing their energies both outwardly and inwardly. Responses to the 93 forced-choice items in the new Form M are all scored to categorize a person as one or the other pole of four, relatively independent bipolar scales: Extraverted-Introverted (E-I), Sensing-iNtuition (S-N), Thinking- Feeling (T-F), and Judging-Perceiving (J-P). Each pole is termed a preference. In Form M, male and female responses to the T-F items are not weighted differently as they are in earlier forms, but still many more women than men score in the Feeling direction. Unlike scores from most other inventories, the scores from the four MBTI scales are grouped into sets of four, termed types. Each preference on a scale is combined with every preference of the other three scales, yielding 16 types. A respondent is assigned the type which has his or her four preferences. Respondents receive their four scale scores and their type score. The manual and accompanying interpretive materials emphasize the type interpretations but 1
provide considerable information about the meanings of individual preference scores. Type descriptions are summaries of key, positively-worded elements from the scales comprising them. Theoretically, scales in a type interact and that interaction modifies their meaning. In addition, the interpretation of the S-N and T-F scales depend in part on whether they are favored by the person in dealing with the external world, that is, whether they are or not dominant. As a consequence, the key descriptors of a scale vary somewhat over the four types of which it is an element. For example, thinking (T) is described as practical and analytical, focused on facts when paired with sensing (S) and as logical and analytical, focused on possibilities when paired with intuition (N). Uses in Counseling Designed for use in training and counseling, the MBTI is one of the most frequently used personality inventories, selling more than two million copies a year. Counselors report that its positive, everyday scale and type descriptions help them to engage clients in affirming self-understanding, exploring educational and career choices, and improving communication at work and in relationships. The manual provides many suggestions from practice of how knowledge of preferences might aid clients in such areas. The authors repeatedly remind users that the MBTI results are descriptions; not prescriptions. Preference scores are intended to suggest factors to consider in decision making and interactions with others. Despite substantial improvements in the MBTI, users need to remember that many of the suggested 2
uses, although more explicit and better illustrated, have not been tested experimentally. As such those suggestions need to be recognized as hypotheses to be tested. This is especially the case if the MBTI is to be substituted for other measures for purposes such as selecting college majors or occupations or improving communication. For example, in a comparison of the ability of interest inventory and MBTI scores to distinguish the subfield preferences of MBA students, Martin and Bartol (1986) found the interest inventory but not the MBTI scores to be effective. Relatedly, in a comparison of the ability of the S-N scores and relationship resource scores to account for differences in the quality of couples communication, Kobes and Lichtenberg (1997) found that relationship resource scores, but not S-N preferences, accounted for differences. The manual emphasizes the importance of clients agreement about their MBTI scores in its interpretation. It provides guides to use in verifying client agreement and suggests that most respondents agree with their four preferences and a majority with their type. One can expect clients to give more credence to suggestions emanating from preferences with which they agree. As a consequence, counselors will want to secure independent judgments from clients. Also new Preference Clarity Categories of slight, moderate, clear and very clear for each preference based on the proximity of a score to its scale s midpoint indicate the portion of a preference s items chosen by the respondent. Slight and moderate preferences are more likely than the clear and very clear preferences to change upon retest. When either preference 3
agreement or clarity is low, the authors appropriately advise caution in interpretation. While positive aspects of each preference are emphasized, the authors allow that there may be deficits associated with each preference and list typical work stressors for each. Findings by Healy & Woodward (1998) that particular MBTI preferences and their interactions relate to some career development obstacles support a preference-deficit linkage and suggest exploring the typical stressors of a client s preferences to identify possible counseling needs. Technical Considerations The MBTI has stirred substantial controversial. Researchers question such features as interpreting individual scales as bipolar categories rather than as a continuous dimension, modifying scale interpretations based upon the type of which they are a member, and also modifying interpretation depending upon whether S, N, T, or F is deemed to be a dominant function. And, while research indicates that the MBTI samples the domains of four of the Big Five Personality Factors, McCrae and Costa (1989) contend that MBTI scale descriptions need to be changed to be consistent with the relationships which MBTI scales have shown with scales of other personality measures. The revised MBTI Form M and the third edition of its Manual are likely to encourage continued use of the MBTI and lessen, but not eliminate, criticisms. The manual provides a clear description of the rationale guiding item selection and summarizes findings from a confirmatory factor analysis which support a four 4
factor solution corresponding to the four MBTI s scales. The manual reports that Form M scores correlate very highly with corresponding scale scores of the earlier Form G. For a sample of 101 adults who completed Forms M and G at the same time, scale correlations ranged from.87 to.90, percentage of agreement on preferences ranged from 79% to 90%, and there was 60% agreement on type classification. The Item Response Theory (IRT) which guided item selection increased scale reliability and maximized discrimination near the midpoints of the four scales. Internal consistency estimates for the National sample of men and women combined were high, ranging from.90 to.92. The manual also reported high test-retest coefficients for the scale scores from three samples, ranging from.83 to.95, and 55% to 80% of the people in these samples were categorized the same type on retest. Comparisons of scores from Form M and G, moreover, indicate that Form M is less likely to yield a preference or type score with which a respondent disagrees. Nevertheless, the controversy over whether the scales are measuring bipolar or continuous variables is likely to persist. In part, this is because the validity of the preferences is based upon their relation to other personality traits which have normal distributions instead of bimodal distributions, and the evidence for those relationships comes from studies which almost always employ continuous MBTI scores. Especially pertinent because of past criticisms are summaries in the manual of results of multiple tests of the claim that types are more than the sums of their parts. Essentially, these studies show that in many instances considering some hypothesized scale interactions increases predicted variance in particular criteria 5
beyond what would be accounted for if one considered only the effects of the scales. Also in some, but not all instances, considering whether the S, N, T, or F functions are dominant increases variance accounted for in particular criteria. While these findings do not always support type interpretations, they strengthen the argument for using type interpretations. The willingness of the authors to undertake such tests on their standardization sample and to share their findings is certainly commendable. Computer Based Version Form M can be taken and scored on the computer using software licensed by the publisher. Booklets can be either computer or hand scored. Computer scoring of Form M yields IRT-weighted scores; hand scoring yields scores that closely approximate them. The authors report very high degrees of agreement between computer and hand scoring for the eight preference scores, ranging from 94% to 99%. Agreement on type classification is not reported. Continuous MBTI scores are now available, however, only through computer scoring. Overall Critique In summary, the 1998 MBTI provides clear, reliable information about a person s preferences that many counselors and clients feel has been helpful. Improvements in the 1998 edition increase the likelihood that it will yield meaningful information. With proper interpretation it can provide useful hypotheses for counselors and their clients to evaluate. As long as counselors 6
are aware of the criticisms cited in the literature, the MBTI is likely to be a useful career counseling tool. References Healy, C.C. & Woodward, G. A. (1998). The Myers-Briggs Type Indicator and career obstacles. Measurement and Evaluation in Counseling and Development, 31, 74-85. Jung, C.G. (1971). Psychological types (H.G. Baynes, Trans., revised by R.F.C. Hall), Princeton: Princeton University Press. (Original work published 1923.) Kobes, K. J. & Lichtenberg, J.W. (1997). Personality similarity, interpersonal perception, and relationship satisfaction. Paper presented at the annual meeting of the American Education Research Association, Chicago, IL. Martin, D. C. & Bartol, K. M. (1986). Holland s Vocational Preference Inventory and the Myers-Briggs Type Indicator as predictors of vocational choice among Master s of Business Administration. Journal of Vocational Behavior, 29, 51-65. McCrae, R. R. & Costa, P. T. (1989). Reinterpreting the Myers-Briggs Type Indicator from the perspective of the five-factor model of personality. Journal of Personality, 57, 17-40. 7